Re: weighting depth clues

[T3D Richard Young <young@xxxxxxx>]

With respect to John Bercovitz's interesting question in TECH-3D 180 on
depth cues, Kent Stevens at the University of Oregon has done extensive
research on the relative importance of various monocular and binocular 
depth cues. I have forwarded your messages and asked for a reply.

Jim Roberts (TECH-3D 181) refers to "Ken Dunkley's pinhole arrangement" 
in which a 3D effect was gotten from 2D shots. One trouble with seeing 
depth in 2D photographs or paintings is that both our bincoular cues 
(stereo vision and convergence) and one monocular cue (accommodation)
tell our brains the image is flat, contradicting the monocular cues for
space perception (relative size, interposition, linear perspective, 
aerial perspective, monocular movement parallal, light and shade). This
contradiction can be corrected by a special eyepiece that sends the
identical image to both eyes (biocular viewing). When 2-D paintings or
photographs are viewed with this device, the depth effects are enormously
increased. (Jan Koenderinck showed me such a device he built when I
recently visited him in the Netherlands).

Jim Crowell (TECH-3D digest 181) states:

> For example, it may be that the people on this
>list tend to apply a relatively high weight to stereo, but the relatively
>high incidence of stereoblindness (7%?) in the general population suggests
>that there isn't a whole lot of evolutionary pressure to use stereo at
>all..

I recently asked my colleague Gerald Westheimer at Berkeley "Why 3D?"
Here is his reply:

>There was an article the other day in the newspaper about a pilot who had an
>accident and it was revealed that, contrary to regulations, he was wearing
>some kind of contact lens correction.  I read that to mean that he was, as
>some vain people do, wearing a contact lens for reading correction in only
>one eye, leaving the other for distance vision.  This seriously impairs
>stereopsis (Suzanne McKee and I wrote a paper on that in JOSA about 1980).
>   As you say, why go to all that trouble and relinquish panoramic vision in
>order to have both eyes with overlapping fields?
>   Incidentally, about 98% of college biology students have acceptable
>stereopsis and 3-D is an exceedingly good tool for demonstrating molecular
>structure.  More people have full stereoptic capability than full color vision!
>
>All good wishes
>
>Gerald


------------------------------